The general public has been made fully aware that cigarette smoking has adverse effects on health and survival prospects. The Surgeon General of the United States concluded that smoking is the one behavior that has been proven to increase mortality chances, especially for certain caused of death. A substantial research literature, particularly in the medical and epidemiological fields, has emerged that documents the basic ties between smoking and life chances. Yet, we lack sufficient knowledge about the demographic and socioeconomic forces and other social contextual factors that increase or decrease the mortal risks of smokers. The literature in this area is sparse and focuses largely on certain populations and particular causes of death. Moreover, it fails to take account simultaneously of a range of factors that influence the smoking- mortality relationships. In the proposed research, sociodemographic conceptual and methodological approaches will be used to examine a model that elucidates the connections between demographic and socioeconomic factors, and other risk behaviors, as they specify the relationship of smoking and mortality. Special emphasis will be given to the interaction of variables and to the varying cause-of-death outcomes. A unique data set has been devised that incorporates files from the 1986 National Mortality Followback Survey, the 1985 National Health Interview Survey, Health Promotion and Disease Prevention Supplement, and the 1987 National Health Interview Survey, Cancer Risk Factor Supplement, Epidemiology Study. Analysis will be done using multinomial logistic regression and log-linear techniques. The results will enable us to extend our knowledge of the complex ways in which population characteristics and other behaviors of individuals modify the mortality outcomes associated with smoking. As a consequence, we should be able to clarify the independent and interactive role of smoking as a determinant of mortality, as well as show the implications of the patterns for the future health and longevity of different subpopulations in the United States.